Microstructural prediction of cast A356 alloy as a function of cooling rate

Abstract

In this paper, a relatively new approach is presented in order to predict the microstructure of A356 using finite element technique and artificial neural network. In the training and test modules of the neural network, different primary and secondary dendrite arm spacing obtained from finite element method were used as inputs and eutectic volume percentage, silicon volume percentage, silicon rod spacing, average length of silicon rods and silicon rod diameter were used as outputs. After the training set was prepared, the neural network was trained using different training algorithms, hidden layers and neurons number in hidden layers. The results of this research were also used to form analytical equations followed with solidification codes for SUT Cast software.